Skip to main content

Catalog Intelligence

Public beta

CORPUS is being built in the open. Some of what you read here is live, some is still design intent — expect it to evolve.

An illustration of a DJ scratching vinyl records on a turntable in the foreground, with two suited men in hats and sunglasses operating a sampler in the background and a CORPUS logo on the right.

Catalog Intelligence is the CORPUS annotation and search stack, applied to your catalogue. We run the same semantic pipeline that makes the CORPUS corpus operational over your tracks, and we give you the resulting index back as a searchable interface for your own teams or as an API for your products.

It is the most direct way to get value from CORPUS's annotation technology today, without waiting for the broader model-licensing infrastructure to mature.

What you get​

The offer comes in two parts:

  • Setup: annotation of your catalogue. A per-track fee covers running the pipeline across your library — extracting acoustic, semantic, narrative, and originality dimensions per track. You receive the full annotation index over your catalogue. See What CORPUS Knows About Each Track for the dimensions captured.
  • Recurring: the search API. Your teams and products query the annotated index through the same three search modes that power our demo — precision, audio similarity, and narrative (Story2Music). The recurring tariff covers usage of the API.

The setup is bookable now. The recurring search API is built to production specification when concrete engagements are in flight; the integration timeline is part of the per-engagement scoping conversation.

Who this is for​

Two customer profiles dominate:

  • Catalogue owners — publishers, libraries, production-music catalogues, sync agencies — typically in the 50,000 to 500,000 track range, who want their catalogue annotated to a higher resolution than today's tagging systems offer and searchable through an interface that matches how creative teams actually look for music.
  • Platforms that embed music — sync platforms, app-builders for situational music, distribution and discovery products — that want a semantic layer they can run their own catalogue against, without having to build the annotation pipeline themselves.

How it differs from existing tagging systems​

In track-by-track benchmarks against Cyanite, Bridge.audio, and AIMS, CORPUS's description quality leads. That is the differentiator: not a different category of feature, but materially better resolution on the dimensions that determine whether a search returns the right track. The underlying reason — descriptions instead of categorical tags — is set out at An Objective Description of Music in the journal.

The other distinguishing piece is the narrative mode. Existing systems tag tracks; CORPUS interprets scenes. A creative team that gives Story2Music a screenplay beat gets back three dramaturgically distinct musical directions, each with retrieval from the catalogue underneath. That capability does not exist in the standard tagging stack.

Evaluating on your own catalogue​

The most direct way to see the quality gap against your current tagging system is on your own material — we are happy to scope a focused test on a slice of your catalogue and walk through the results with you. Terms depend on the size and shape of the test; the conversation starts in Getting Started.